Apple explores AI chip acquisitions to boost server hardware

Apple is actively exploring acquisitions of artificial intelligence chip companies to accelerate the development of its own server processors, a strategic shift aimed at reducing its reliance on external hardware for demanding AI workloads. The company has reportedly approached semiconductor startups and consulted with bankers about potential deals, signaling a new willingness to make larger purchases to close a significant performance gap in its AI infrastructure.

Currently, Apple handles some AI processing in its own data centers using its internally designed M2 Ultra chips, which power its Private Cloud Compute servers. However, these chips have proven insufficient for more complex tasks. A concrete example emerged during development of a major Siri overhaul, which incorporates Google's Gemini AI models. When engineers attempted to run these advanced models on Apple's own servers, the M-series hardware could not handle the load. As a result, more demanding Siri tasks are being routed through Nvidia graphics processing units hosted within Google's cloud infrastructure.

This dependency conflicts with Apple's longstanding preference for controlling its own silicon and...